Fraunhofer Center for Silicon Photovoltaics CSP
Fraunhofer Center for Silicon Photovoltaics CSP

PV System Diagnostics

Services

  • Physical-technical, empirical or machine learning-based system modeling and analysis of yield losses
  • Rapid identification of defect patterns in operational data
  • Simulation of different operating scenarios using models of system-specific performance
PV plant with snow
© Fraunhofer CSP
Snow and other weather conditions can lead to energy yield losses of PV systems.

Failure Analysis for Facilitated Plant Monitoring
 

  • Identification of operating conditions
  • Verification of alarms
  • Determination of energy yield loss
  • Energy yield prediction
Pyranometer
© Fraunhofer CSP

Risk Minimization
 

  • Improvement of data quality
  • Digitalization of PV systems
  • Sensor concepts for innovative data-driven applications
Module Soiling
The soiling of modules is a key factor in reducing maintenance costs.

Reducing Operation and Maintenance Costs
 

  • Determination of site-specific energy yields
  • Energy yield simulation of different maintenance scenarios
  • Predictive maintenance
Elektrolumineszenz-Aufnahme PID
© Fraunhofer CSP
Electroluminescence image of a solar module affected by PID.

Damage Case Assessments
 

  • Defect detection
  • Impact of defects on total energy yield
  • Impact of module aging on long-term energy yield planning

Credentials

Promotional poster VR4PV
EFRE funding logo Saxony-Anhalt
EFRE funding logo Saxony-Anhalt

Project VR4PV

  • Creation of a virtual environment and a digital image of PV systems for future analysis, inspection and maintenance (VR4PV)
  • Joint project of Fraunhofer CSP (Halle), Fraunhofer IFF (Magdeburg), DENKweit GmbH (Halle), Dexor Technology GmbH (Köthen)

Duration

January 04, 2022 - December 31, 2022


Project Goals

  • Application of suitable imaging methods for the acquisition of the PV system as a whole as well as on component level in combination with the necessary geolocation
  • Development and application of data routines for automated "recognition" and "assignment" of system-relevant variables and correlations using Deep Learning
  • Development of a data management concept for structuring a database of the acquired and modeled data (health record)

Head of Project

Dr. Matthias Ebert – matthias.ebert@csp.fraunhofer.de